Comments on: Longitudinal vs Cross-sectional http://ozrisk.net/2008/03/27/longitudinal-vs-cross-sectional/ Risk Management in Australia Wed, 30 Nov 2011 12:41:58 +0000 hourly 1 http://wordpress.com/ By: Clive http://ozrisk.net/2008/03/27/longitudinal-vs-cross-sectional/#comment-26229 Wed, 09 Apr 2008 15:30:11 +0000 http://ozrisk.wordpress.com/?p=307#comment-26229 In press articles it is common to see a statistic like “Home loan default rate this month is 1%”, by which they presumably mean that, this month, there were 654,321 HLs of which 6,543 were in default. Alternatively, the ratio might be worked in dollar values rather than numbers of accounts. These would be the classic cross-sectional only statistics, and much of any bank’s frontline reporting would be of this type: for example, percentages in current, 30DPD, 60DPD, 90DPD,..

This is good enough for the big picture if one is not worried about effects due to changes in seasoning of the portfolio, or to changes in collection procedures, or differences between different cohorts (vintages).

Not sure if you had some more specific question in mind?

]]> By: Chris http://ozrisk.net/2008/03/27/longitudinal-vs-cross-sectional/#comment-26228 Wed, 09 Apr 2008 07:46:06 +0000 http://ozrisk.wordpress.com/?p=307#comment-26228 Can you point out some examples of the cross-sectional only risk monitors?

]]>